I'm on a call right now with an ICU physician who is working bioengineers at his academic hospital to jury-rig extra ventilators. He has a lot to say about the various approaches, including Medtronic's contribution (basically not useful).

A brief thread. https://twitter.com/SebastianSeung/status/1247522200892121088

This guy just got off ventilating an 18-year-old, previously healthy, who was felled by coronavirus. This is not just a disease of old people.

He has seen some online designs for venting multiple patients, basically by putting the patients in series. His verdict: "dumb as shit."

The basic issue is that parameters need to be adjusted for each patient individually. Respiratory therapists & doctors need both pressure and flow measurements.

Venting multiple patients can be done, but it is more complicated. A key bottleneck is flow sensors.

There is one ventilator design out there that works, from Matt Maltese at Pitt. Includes a flow sensor.

(There are local efforts to produce flow sensors in large numbers. To learn more about that, join @SebastianSeung's discord channel. event at 5pm today)

Other approaches will not be useful:

1) Medtronic donated design - it's an old plan, will never be produced, and certainly not in time.

2) @ElonMusk's "donation" - those are BiPAPs. Worthless. Irresponsible and probably to pump up image and maybe stock price.

Now, practicalities of being in the southward advancing wave of coronavirus. In the central NJ/Phila. area, we're getting "medical refugees." Basically people fleeing the advancing wave.

Hospitals understaffed. Interns, residents. Not experts - and they need pressure/flow data.

It sounds like respiratory therapists will play a key role. Doctors make "decisions" but the therapists can manage up, i.e. tell the doctor what to do. So the intern/resident thing might be OK, if the respiratory therapists get a good way to monitor lots of patients at once.

Many approaches to making emergency ventilators are springing up around the U.S. Whose approach will work? Which is best?

This guy says don't worry about that. Massive surge coming, need to prepare by whatever means. And another wave later maybe. Let a hundred flowers bloom.

For flow meters specifically, "There is no large corporation coming around the corner to solve these problems. They will be mired in whatever stuff they have to do. Nothing is coming that we can buy."

There are multiple teams around the country trying to solve these problems. Some "Cadillac" approaches, some simpler.

I am now watching scientists and engineers in my community work on this problem. It is an inspiration to see them come together. One flower blooming here in N.J.